Dokumen.tips_the-impact-of-ownership-structure-on-corporate-debt-policy-a-time-series-cross-sectional.docx

EFA Eastern Finance Association

The Financial Review The Financial Review 33 (1998) 85-98

The Impact of Ownership Structure On Corporate Debt Policy: a Time-Series Cross-Sectional Analysis Mahmoud A. Moh'd Saint Mary's University

Larry G. Perry University of Arkansas

James N. Rimbey* University of Arkansas

Abstract This study examines the influence of agency costs and ownership concentration on the capital structure of the firm. Of particular interest is the composition of equity ownership as a determinant of overall capital structure and the dynamic adjustment of capital structure to changes in the equity ownership. Results indicate that the distribution of equity ownership is important in explaining overall capital structure and that managers do reduce the level of debt as their own wealth is increasingly tied to the firm. It is also noted that the time-series component is important in resolving the conflicting results reported in prior research. Keywords: capital structure; agency theory; equity ownership distribution JEL classification: G32/G34

1. Introduction Agency theory has been advanced as a possible explanation for the observed variation in capital structures across firms. Jensen and Meckling (1976) suggest that

*Corresponding author: Department of Finance BA-302, College of Business Administration, University of Arkansas.Fayetteville AR 72701, Phone: (501) 575-5259, Fax: (501) 575-8407,E-mail:jrimbey@com• p.uark.edu.

85

'd, L.G. Perry, J.N. Rimbey/The Financial Review 3333 (1998) 85-98 8686 M.A. M.AMoh . Moh'd, L.G. Perry, J.N . Rimbey/The Financial Review (1998) 8598 8686

a manager's capital structure decision amounts to balancing the agency cost of debt against the agency cost of equity to minimize their impact on the value of the firm. This paper attempts to resolve some of the uncertainty surrounding the question of ownership structure and corporate debt policy. An extensive time-series cross• sectional analysis is employed to examine the dynamic response of capital structure to agency variables through time within firms, as well as across firms. This adds additional information and allows stronger conclusions than the conflicting results derived from use of static cross-sectional models reported to date. The agency theory articulated by Jensen and Meckling posits that there exists a natural conflict in the interests of outside shareholders and the managers of a firm, leading to the possibility that managers may make suboptimal decisions that improve their own welfare at the expense of shareholders. Recognizing the impact of these conflicts between owners and managers, the market makes unbiased estimates of such costs and reduces the value of firm shares accordingly. This loss is the firm's agency cost of equity. Several methods have been suggested to mitigate the agency problem and reduce the attendant costs. These methods fall into the two categories of either external control devices or motivational mechanisms. In the latter group, one way of bringing the interests of owners and managers into closer alignment is to cause managers to increase their ownership in the firm (Jensen and Meckling, 1976). A second way suggested by Jensen and Meckling is to increase the use of debt financing, in effect displacing equity capital. This shrinks the equity base, thus increasing the percentage of equity owned by management. The use of debt not only serves to align the interests of managers and owners, but also increases the probability of bankruptcy and job loss. This added risk may further motivate manag• ers to decrease their consumption of perks and increase their efficiency (Grossman and Hart, 1982). Another benefit of the use of debt is a solution to the problem of the over-retention of earnings (Jensen, 1986). While the outcome of each of these predictions may appear to be clear, the empirical evidence is not. Kim and Sorensen (1986), Agrawal and Mandelker (1987) and Mehran (1992) find a positive relationship between the percentage of shares held by insiders and the firm's debt ratio. Using different data sources, sample periods and sampling techniques, Friend and Hasbrouk (1988), Friend and Lang (1988) and Jensen, Solberg, and Zorn (1992) report a negative relationship. In a study requiring pre-set levels of share ownership, Chaganti and Damanpour (1991) find that insider stock ownership has no impact on the capital structure of the firm. Amid this uncertainty there is yet another aspect to be considered in the ownership question. The nature of the distribution of shares among the outside shareholders has been suggested as a device to mitigate agency costs, thus affecting the firm's capital structure. Since ownership represents a source of power that can be used either to support or oppose existing management, the concentration or dispersion of that power becomes relevant. While prior research has viewed the effect of agency variables on certain characteristics of the firm, this study examines

'd, L.G. Perry, J.N. Rimbey/The Financial Review 3333 (1998) 85-98 8787 M.A. M.AMoh . Moh'd, L.G. Perry, J.N . Rimbey/The Financial Review (1998) 8598 8787

the equity ownership structure as it affects corporate debt policy and improves upon prior research in several ways.

1. Tests are conducted using time-series cross-sectional analysis across 18 years to measure the response of debt policy changes in specific agency variables through time as well as across firms. 2. In a second set of tests, the variables are entered in first difference form to capture the dynamic nature of firm behavior. 3. Measures of operating risk and pure business risk are explicitly included in the model rather than allowing one variable (the firm's beta) to proxy for both. 4. A set of refined explanatory variables, based in theory, are included in the model.

2. Variable definition and model design A model is constructed to test for the effect of ownership structure on corporate debt policy. The dependent variable is the firm's debt ratio, defined as the book value of long term debt divided by the sum of the book value of long term debt plus the market value of equity. The determinants of capital structure include explanatory variables representing the various agency costs as follows.

2.1. Ownership structure In his study of agency costs and the dividend paying behavior of firms, Rozeff (1982) argues that dividend payments are a part of the firm's monitoring/bonding package, in that firms tend to pay out more in dividends when insiders own a lower proportion of shares. As in Rozeff, the variable representing inside ownership in this model is the percentage of shares owned by insiders as reported by Value Line Investment Survey. Two measures of outside shareholder concentration are also considered in the ownership structure for the monitoring role they might play. The first is the natural log of the number of outstanding shareholders as reported by Value Line. Rozeff argues that the greater the number of shareholders, the more diffused the ownership, hence a negative or insignificant relationship should be expected between the number of shareholders and the level of debt. The second proxy represents the make-up of outside ownership and is included to examine the Grier and Zychowicz (1994) tenet that institutional investors may substitute for the disciplinary role of debt in the capital structure. This variable is defined as the percentage of shares owned by institutional investors, also as reported by Value Line.

'd, L.G. Perry, J.N. Rimbey/The Financial Review 3333 (1998) 85-98 8888 M.A. M.AMoh . Moh'd, L.G. Perry, J.N . Rimbey/The Financial Review (1998) 8598 8888

2.2. Dividend payments Agency theorists have drawn a link between the issuance of debt and the payment of cash dividends (e.g., Jensen, Solberg, and Zorn, 1992). Specifically, it is suggested that dividend payments and debt act as substitutes in reducing agency costs. For this reason, dividend payout serves as an explanatory variable with an hypothesized inverse relationship.

2.3. Growth opportunities Myers (1977) identifies growth opportunities as a significant determinant of capital structure, for which Rozeff (1982) utilizes Value Line's forecast of five-year sales growth as proxy. Since this figure represents an ex ante estimate of growth opportunities, it is also used in the model.

2.4. Firm size Numerous studies have argued that the debt policy of firms may be affected by size, each suggesting a positive relationship between the firm's size and debt ratio. Following Titman and Wessels (1988), the natural log of sales is employed as proxy for size.

2.5. Asset structure Prior research suggests that the composition or collateral value of the firm's assets influence its financing sources (e.g., Myers and Majluf, 1984). Therefore, there should be a positive relationship between the level of debt and the collateral value of assets. Following Titman and Wessels (1988) and Mehran (1992) the ratio of inventory plus gross plant and equipment to total assets is used as proxy for the firm's asset structure.

2. 6. Asset risk Long and Malitz (1985) argue that the firm's systematic risk (beta) captures the firm's business risk in addition to its financial risk, and they follow the procedure suggested by Hamada (1972) to unlever beta. However, Mandelker and Rhee (1984) and Rhee (1986) argue that beta is composed of three levels of risk: intrinsic business risk, financial leverage risk, and operating leverage risk. While this distinction may seem subtle, due to the fact that these risks can offset one another it is important that they be considered in any examination of the agency costs of debt.

'd, L.G. Perry, J.N. Rimbey/The Financial Review 3333 (1998) 85-98 8989 M.A. M.AMoh . Moh'd, L.G. Perry, J.N . Rimbey/The Financial Review (1998) 8598 8989

In this study beta is decomposed to isolate intrinsic business risk and operating leverage risk as outlined in Rhee's model. 1 The firm's beta is calculated using the standard form of the market model and Center for Research in Security Prices (CRSP) daily returns over each full year. Book value of debt proxies for market value of debt (Bowman, 1980), and the balance sheet items specified in Rhee's model are derived from Compustat.

2. 7. Profitability Profitability reflects earnings to finance investment. Myers (1984) suggests managers have a pecking order in which retained earnings represents the first choice, followed by debt financing, then equity. If this were true it would imply a negative relationship between profitability and the debt ratio. A number of prior studies define profitability as the ratio of operating income to total assets, which is the proxy employed in this model.

2.8. Tax rate With the relaxation of irrelevance assumptions, firms with high tax liabilities are expected to utilize greater amounts of debt to take advantage of the deductibility of interest expense (e.g., Haugen and Senbet, 1986). To account for this, Zimmerman's (1983) ratio of taxes paid to pre-tax income is employed as the tax rate proxy.

2.9. Non-debt tax shield DeAngelo and Masulis (1980) argue that non-debt tax deductions substitute for the tax shield benefits of debt. As a result, firms with greater non-debt tax shields would be expected to have lower levels of debt. In this study non-debt tax shield is defined as in Titman and Wessels (1988), the ratio of depreciation, invest• ment tax credit, and tax loss carryforward to total assets.

I

Following Rhee (1986): operating leverage risk = (1 - t)j3° CIS; financial leverage risk (1 - t) 13° DIS; intrinsic business risk = j3°;

=

where: t = corporate income tax rate; 13 = systematic risk; C risk-adjusted present value of total fixed costs; D = market value of risky debt; S = market value of common equity; and 13° = W[l - t)(CIS + DIS) + 1].

=

90

M.A. Moh'd, N. Rimbey/The Moh 'd, L.G. Perry, Perry, J. J.N. Rimbey/The Financial Review 33 (1998) ( 1998) 85-98

9090

2.10. Uniqueness Titman and Wessels (1988) also argue that the capital structure decision takes into account the risk of bankruptcy. The more unique the firm, the higher the risk, thus the higher the cost of bankruptcy. Hence, a negative relationship is expected between the firm's degree of uniqueness and its debt ratio. The Titman and Wessels measure for uniqueness, defined as the ratio of advertising plus research and develop• ment expense to annual sales, is the variable included in this model. A basic premise of this study is that the variability in the debt ratio is a function of the firm's dynamic behavior through time. Therefore, a time-series cross-sectional approach (TSCS) is employed to analyze the data and examine the impact of agency costs on the debt policy of firms. TSCS contains the necessary mechanism to deal with both the intertemporal dynamics and the individuality of the firms being investigated. The methodological improvements gained from pooling TSCS data are well documented (e.g., Judge, Griffiths, Hill, and Lee, 1980; Dielman, 1983; and Hsiao, 1986). Also reported are the results of the individual cross-sectional and time-series components of the full model. By definition, the cross-sectional regression uses the average value for each variable, thus ignoring changes through time. It is a measure of the agency effects on debt policy across firms. The time-series portion is a ''fixed effects" construct which examines the agency effects on debt policy on the basis of within-industry variation through time, ignoring cross-sectional changes between firms. The use of the industry level grouping based on two-digit SIC codes is employed to reduce measurement error in the variables as explained in Smith and Watts (1992). Including industry dummy variables in the model relaxes the assump• tion of a common intercept, allowing the intercept to vary as a means of representing individual industry effects (Hsiao, 1986). Because many of the agency proxy vari• ables have been subject to mixed results in previous research, two-tailed t-tests are used to test the significance of the regression coefficients for each variable. The model is expressed as follows: Debt ratio = f (percent insider ownership, percent institutional ownership, shareholder dispersion, cash dividends, firm growth, firm size, asset structure, intrinsic business risk, operating leverage risk, firm profit• ability, tax rate, non-debt tax shield, uniqueness of the firm). To measure dynamic behavior through time, it is assumed that firms evaluate their financial position periodically and adjust their debt ratios based on anticipated internal and external forces. This can be tested in two ways. The first is a TSCS test with an appropriate lag between the explanatory variables and the debt ratio; i.e., entering the prior period's debt ratio as an independent variable in accordance with the dynamic model construction (Hsiao, 1986). This is accomplished in a series of tests which include the difference in the year-to-year firm debt ratios on the left side of the equation, with lagged debt on the right. As explained later, this allows measurement of the annual rate of adjustment. The second way of testing the adjustment of firm debt to change in the agency variables is to model the difference

91

M.A. Moh'd, N. Rimbey/The Moh 'd, L.G. Perry, Perry, J. J.N. Rimbey/The Financial Review 33 (1998) ( 1998) 85-98

9191

in year-to-year firm debt as a function of the year-to-year changes in the explanatory variables. This study reports the results of both methods. The sample of 311 firms is obtained by examining annual financial statement data over the period 1972-1989 as reported in the Compustat Industrial File. The study includes only those unregulated manufacturing firms for which complete data for all variables is available. Firms so identified also must be listed in the CRSP Daily Master File and the Value Line Investment Survey over the entire 18-year period.

3. Results Descriptive statistics for the variables included in the model are presented in Table 1. The mean debt ratio across firms through time is 23.4%.2 Insider holdings range from 0% to 72%, with a mean of 14.8%. Likewise, institutional holdings and shareholder dispersion vary widely, offering a large continuum along which to measure the hypothesized agency effects. Diagnostics show no severe multicollinear• ity between variables. First-order autocorrelation and heteroskedasticity are present, however application of the Parks (1967) GLS transformation alters neither the sign nor significance of the model parameters. Hence, the OLS results representing the interpretable impact of the explanatory variables without transformation are reported here. The first column of Table 2 presents the results of the primary TSCS analysis, with debt as the dependent variable. The coefficients of all explanatory variables are significant at the 0.0001 level. The adjusted R2 of the model is 0.464. These results provide support for the proposition that the level of debt in the capital structure is inversely related to insiders' shareholdings, which is consistent with the various theoretical considerations (e.g., Easterbrook, 1984). The institutional shareholdings variable is significantly negative, supporting the argument that institu• tional investors may substitute for the disciplinary role of debt in the capital structure. The significantly negative impact associated with the number of outside stock• holders supports the argument that diffused shareholders have little effect or influence on management's conservative position on debt issuance (e.g., Easterbrook, 1984). In summary, the evidence on the ownership structure of equity is consistent with the argument that higher insider ownership accompanied by diffuse outside shareholdings permits managers to control financial policies of the firm and pursue their own interests, as posited by agency theory. Interpretation of the results of the effects of the other related variables included in the model also have meaning and help in understanding the theoretical considerations.

2

While macroeconomic conditions did cause fluctuations in the capital structure of firms over time, there is no evidence of any structural shift in overall debt ratios during the period under study.

92

M.A. Moh'd, N. Rimbey/The Moh 'd, L.G. Perry, Perry, J. J.N. Rimbey/The Financial Review 33 (1998) ( 1998) 85-98

9292

Table 1

Summary statistics Firm characteristics relating to ownership structure and corporate debt policy for 311 firms over the years 1972-1989.

Variable

Mean

Std. Dev.

DEBT INSDR INSTINV STKHLDR DIVIDEND GROWTH SIZE ASSET BUSRISK OLRISK PROFIT TAXRATE NDTXSHD UNIQUE

0.2344 0.1484 0.2978 9.5376 0.4003 0.0945 6.8942 0.5989 0.7186 0.2167 0.1664 0.4019 0.0580 0.0318

0.1874 0.1604 0.2141 1.2609 0.2676 0.0455 1.4813 0.1501 0.3574 0.1493 0.0758 0.1628 0.0813 0.0409

Minimum

Maximum

0.0000 0.0010 0.0000 6.9078 0.0000 -0.1100 3.1800 0.0523 0.0100 0.0200 -0.2635 0.0000 0.0140 0.0000

0.9436 0.7200 0.8592 14.0954 1.0000 0.4100 11.7360 0.9602 3.1496 1.4615 0.4697 1.0000 1.7736 0.3446

DEBT = book value long term debt to book value long term debt plus market value equity INSDR = percentage of common stock held by insiders INSTINV = percentage of common stock held by institutional investors STKHLDR = natural log of number of shareholders DIVIDEND = dividend payout ratio GROWTH= a forecast of future five-year sales growth SIZE = natural log of firm sales ASSET = ratio of inventory plus gross plant and equipment to total assets BUSRISK = intrinsic business risk OLRISK = operating leverage risk PROFIT = ratio of operating income to total assets TAXRATE = ratio of taxes paid to pre-tax income NDTXSHD = ratio of depreciation, investment tax credit, and tax loss carryforward to total assets UNIQUE = ratio of advertising and R&D expenses to sales

As suggested by Jensen, Solberg, and Zorn (1992), dividend payments appear to act as a substitute for debt in the capital structure, serving to cause managers to reduce agency costs. Growth has a significantly negative coefficient, lending support to the Myers (1977) argument that firms with higher investment in intangible assets tend to use less debt in their capital structure to reduce the agency costs associated with risky debt. Size is significant with a positive sign, indicating that firms tend to increase leverage as they grow larger.

9393 M.A. M.A.Moh/d, Moh 'd,L.G. L.G.Perry, Perry,J.N. J.N.Rimbey/The Rimbey/The FinancialReview Review3333 ( 1998)85-98 85Financial (1998) 98

9393

Table 2 Regression results Time-series cross-sectional regression of corporate debt structure on variables impacting the equity ownership structure of 311 firms during the years 1972-1989. The variables are as defined in Table 1.

Independent Variables

Pooled TSCS OLS Regression

CrossSectional Regression Alone

Time-Series Regression Alone

Intercept

0.3444 (15.700)*

0.5691 (6.309)*

INSDR

-0.0953 (-6.889)*

-0.1532 (-2.919)*

-0.0521 (-3.813)*

INSTINV

-0.1460 (-11.851)*

-0.2092 (-2.762)*

-0.1483 (-12.423)*

STKHLDR

-0.0083 (-2.871)*

-0.0181 (-1.630)***

-0.0129 (-4.313)*

DIVIDEND

-0.0500 (-6.152)*

-0.1456 (-2.775)*

-0.0348 (-4.550)*

GROWTH

-0.3370 (-7.247)*

-1.2322 (-4.059)*

-0.2245 (-5.052)*

NIA

0.0184 (7.145)*

0.0216 (2.260)**

0.0246 (9.023)*

0.1849 (14.290)*

0.2609 (5.738)*

0.1534 (10.569)*

BUSRISK

-0.1286 (-22.292)*

-0.0811 (-2.440)**

-0.1395 (-24.698)*

OLRISK

0.2514 (17.800)*

0.1012 (1.491)

0.3291 (23.319)*

PROAT

-0.9863 (-32.793)*

-1.0766 (-7.842)*

-0.9065 (-30.765)*

TAXRATE

0.1554 (12.054)*

0.0207 (0.252)

0.1999 (15.864)*

NDTXSHD

0.1363 (5.858)*

SIZE ASSET

0.2806 (2.006)**

0.1322 (6.003)*

UNIQUE

-0.8496 (-16.912)*

-0.5196 (-2.959)*

-0.8706 (-15.736)*

F-Statistic Adj. R2

374.378* 0.4644

35.538* 0.5916

168.674* 0.5324

* Significant at the 1% level. ** Significant at the 5% level. *** Significant at the 10% level.

9494 M.A. M.A.Moh/d, Moh 'd,L.G. L.G.Perry, Perry,J.N. J.N.Rimbey/The Rimbey/The FinancialReview Review3333 ( 1998)85-98 85Financial (1998) 98

9494

Table 2

The variable associated with collateral value of assets shows a significantly positive relationship with debt. This is in accord with the argument that tangible assets decrease the agency costs of debt (Myers, 1977), the secured debt hypothesis (Scott, 1977), and the pecking-order hypothesis (Myers, 1984). Of particular interest is the fact that firms' debt ratios relate negatively to intrinsic business risk and positively to operating leverage risk. This decomposition of beta into its component parts has more meaning than in prior studies. The results support the theoretical arguments of Myers (1977), Long and Malitz (1985) and Fischer, Heinkel, and Zechner (1989), that the benefits of operating risk in reducing the agency costs of debt and/or providing additional tax benefits may outweigh the bankruptcy risk inherent in high operating leverage. It may also be the case that the positive result gives indirect support to the secured debt hypothesis and other theories related to the asset structure of the firm. Profit has a significantly negative coefficient, which again supports the Myers (1984) pecking-order hypothesis. Tax rate has a positive and significant impact on the debt ratio. Kim and Sorensen (1986) attribute their finding of no significance for the tax rate to the fact that their model may not be dynamic enough to capture the interaction between debt and taxes. The model employed here is constructed specifically to avoid this shortcoming. Non-debt tax shield is positive and significant, which refutes the argument of DeAngelo and Masulis (1980). A possible explanation is that since the primary component of the non-debt tax shield is depreciation, and high depreciation charges may be characteristic to firms with higher levels of fixed assets, then fixed assets may have collateral value and thus increase the debt capacity of the firm. The uniqueness variable is significant and negative. This supports the Titman and Wessels (1988) argument that firms whose high risk of bankruptcy could impact its non-owner stakeholders tend to have lower levels of debt to reduce that risk. The cross-sectional and time-series regression results alone are shown in col• umns 2 and 3 of Table 2. These indicate how much of the effect of the TSCS regression results can be attributed to the differences in debt ratios among firms and how much can be attributed to debt ratio adjustment through time. The adjusted R2 values of 0.59 and 0.53, respectively, indicate that the hypothesized variables explain a substantial portion of both the cross-sectional and time-series variation in firm leverage. The fact that each of these R2 values exceeds that of the TSCS regression (0.464) is also as expected (Bjorklund, 1989). The loss of significance for operating leverage risk and tax rate in the cross-sectional model (Column 2) may be attributable to averaging the data over 18 years rather than allowing it to vary through time. The dynamic adjustment of firm debt ratios to changes in the agency variables can be observed in Tables 3 and 4. Table 3 shows the relationship between dependent and independent variables when the dependent variable is defined as the year-to• year change in debt ratio, and the prior year's debt ratio is entered into the model as an additional independent variable. The purpose is to test the dynamic relationship

L.G. Perry, J.NJ.N. . Rimbey/The Financial Review 3333 (1998) 85-98 9595M.A. M.AMoh'd, . Moh'd, L.G. Perry, Rimbey/The Financial Review (1998) 8598

9595

Table 3

Regression results of first differences in dependent variable, with one-year lag Time-series cross-sectional regression of year-to-year change (first difference) in corporate debt structure on variables impacting the equity ownership structure, including one year lag of debt ratio, to measure the dynamic response through time for 311 firms during the years 1972-1989. The variables are as defined in Table 1. Variable

Parameter Estimate

t-stat

Prob> ltl

Intercept LAGDEBT INS DR INSTINV STKHLDR DIVIDEND GROWTH SIZE ASSET BUSRISK OLRISK PROFIT TAXRATE NDTXSHD UNIQUE

0.0651 -0.2052 -0.0248 -0.0569 -0.0007 0.0223 0.1331 0.0048 0.0488 -0.0447 0.0248 -0.3341 0.0362 0.0024 -0.1006

5.057 -27.859 -3.090 -7.949 -0.396 4.757 4.900 3.222 6.481 -13.101 2.976 -18.042 4.881 0.184 -3.362

0.0001 0.0001 0.0020 0.0001 0.6921 0.0001 0.0001 0.0013 0.0001 0.0001 0.0029 0.0001 0.0001 0.8544 0.0008

Adj. R' = 0.1819

F-Statistic = 84.951

Prob. > F = 0.0001

whereby firms adjust their levels of debt in response to internal and external forces as previously discussed. Results show lagged debt to be highly significant, with the coefficient attaching to the lagged debt ratio being -0.2052. This indicates an implied annual rate of adjustment in the debt ratio to shifts in the other variables of about 21 %, which appears reasonable. One other effect observed with lagged debt is that the variables associated with non-debt tax shield and number of shareholders become insignificant, and the growth and dividend variables switch sign. The now positive impact of growth during the adjustment process may indicate that the adjustment in debt level is partially driven by the need for external financing for new projects, with debt financing receiving priority over equity (Myers, 1984). Another possible explanation is that growth adds value to the firm, which increases borrowing capacity (Titman and Wessels, 1988). The sign change on the dividend variable may relate to the Easterbrook (1984) argument that dividend increases reduce the equity base, thus increasing the debt ratio. As a second test of the influence of changes in the agency variables on adjust• ment of debt ratios, the year-to-year change in debt ratios was regressed on year• to-year changes in the explanatory variables. The results in Table 4 show which variables have greatest impact on the debt ratio adjustment process. While the first differences in the growth variable, collateral value of assets, and non-debt tax shield

L.G. Perry, J.NJ.N. . Rimbey/The Financial Review 3333 (1998) 85-98 9696M.A. M.AMoh'd, . Moh'd, L.G. Perry, Rimbey/The Financial Review (1998) 8598

9696

Table 4

Regression results of first differences in both dependent and independent variables Time-series cross-sectional regression of the year-to-year change (first difference) in corporate debt structure on the year-to-year changes (first differences) in variables impacting the equity ownership structure to measure the dynamic response through time for 311 firms during the years 1972-1989. The variables are as defined in Table 1. Variable

Parameter Estimate

t-stat

Prob> ltl

Intercept AINSDR AINSTINV ASTKHLDR ADIVIDEND A GROWTH ASIZE AASSET ABUSRISK AOLRISK APROFIT ATAXRATE ANDTXSHD AUNIQUE

-0.0001 -0.0092 -0.1390 0.0215 0.0353 0.0433 0.0687 -0.0289 -0.0856 0.1558 -0.4593 0.0997 0.0229 -0.4134

-0.072 -3.566 -9.587 2.717 7.105 1.410 9.342 -1.422 -21.508 13.829 -14.971 12.366 1.036 -2.730

0.9422 0.0004 0.0001 0.0066 0.0001 0.1586 0.0001 0.1550 0.0001 0.0001 0.0001 0.0001 0.3001 0.0063

Adj. R2

= 0.1815

F-Statistic

= 91.179

Prob. > F

= 0.0001

lose significance in this model, the remaining proxy variables are quite robust. The one surprise is the reversal of sign on the proxy representing number of shareholders. One possible explanation is that increases in debt attract investor attention by signaling a third-party review as suggested by Easterbrook (1984). The results of each of these tests support the dynamic nature of adjustment of the firm's capital structure in response to shifts in the ownership structure through time.

4. Conclusion In this paper, the dynamic adjustment of capital structure of firms is examined in an agency theory context. Of particular interest is the ownership structure of the firm's equity as a determinant of overall capital structure, and especially how the capital structure is adjusted in response to shifts in agency cost trade-offs. Studies to date have attempted to test such dynamic relationships using only static models, and have reported mixed results. The results reported here show that the tenets of agency theory appear to hold not only across firms, but also within firms across time. Managers act to adjust the capital structure of firms in response to variations in the agency cost structure in a dynamic manner. Further, it appears the structure of equity ownership is important in explaining the overall capital structure of the firm. As managerial ownership increases, thus raising the amount of personal wealth and human capital invested

L.G. Perry, J.NJ.N. . Rimbey/The Financial Review 3333 (1998) 85-98 9797M.A. M.AMoh'd, . Moh'd, L.G. Perry, Rimbey/The Financial Review (1998) 8598

9797

in the firm, managers tend to lower debt to reduce their overall risk and/or agency costs. Institutional shareholdings also appear to influence the financial policies of firms, with institutional holders substituting for the disciplinary role of debt in the capital structure. When outside ownership is diffuse, those outside shareholders have little influence on managers' conservative debt postures. These are stronger conclusions than the conflicting results derived from use of static cross-sectional models in prior research. The results also suggest that capital structure models that do not include the impact of agency costs or their effect through time may be incomplete.

References Agrawal, A. and G. Mandelker, 1987. Managerial incentives and corporate investment and financing decisions, Journal of Finance 42, 823-837. Bjorklund, A., 1989. Potentials and pitfalls of panel data, European Economic Review 33, 537-546. Bowman, R., 1980. The importance of a market value measurement of debt in assessing leverage, Journal of Accounting Research 18, 242-254. Chaganti, R. and F. Damanpour, 1991. Institutional ownership, capital structure, and firm performance, Strategic Management Journal 7, 479-491. DeAngelo, H. and R. Masulis, 1980. Leverage and dividend irrelevancy under corporate and personal taxation, Journal of Finance 35, 453-464. Dielman, T., 1983. Pooled cross-sectional and time-series data: a survey of current statistical methodology, The American Statistician 37, 111-122. Easterbrook, F., 1984. Two agency-cost explanations of dividends, American Economic Review 74, 650-659. Fischer, E., R. Heinke!, and J. Zechner, 1989. Dynamic capital structure choice: theory and tests, Journal of Finance 44, 19-40. Friend, I. and J. Hasbrouck, 1988. Determinants of capital structure, Research in Finance 7, 1-19. Friend, I. and L. Lang, 1988. An empirical test of the impact of managerial self-interest on corporate capital structure, Journal of Finance 43, 271-281. Grier, P. and E. Zychowicz, 1994. Institutional investors, corporate discipline, and the role of debt, Journal of Economics and Business 46, 1-11. Grossman, S. and 0. Hart, 1982. Corporate financial structure and managerial incentives, in John McCall, ed., The Economics of Information and Uncertainty, (University of Chicago Press, Chicago, IL). Hamada, R., 1972. The effect of the firm's capital structure on the systematic risk of common stocks, Journal of Finance 27, 435-452. Haugen, R. and L. Senbet, 1986. Corporate finance and taxes: a review, Financial Management 15, 5-22. Hsiao, C., 1986. Analysis of Panel Data, (Cambridge University Press, Cambridge). Jensen, M., 1986. Agency costs of free cash flow, corporate finance, and takeovers, American Economic Review 76, 323-329. Jensen, M. and W. Meckling, 1976. Theory of the firm: managerial behavior, agency costs, and ownership structure, Journal of Financial Economics 3, 305-360. Jensen, G., D. Solberg, and T. Zorn, 1992. Simultaneous determination of insider ownership, debt, and dividend policies, Journal of Financial and Quantitative Analysis 27, 247-263. Judge, G., W. Griffiths, R. Hill, and T. Lee, 1980. The Theory and Practice of Econometrics, (John Willey and Sons, New York, NY). Kim, W., and E. Sorensen, 1986. Evidence on the impact of the agency costs of debt on corporate debt policy, Journal of Financial and Quantitative Analysis 21, 131-144. Long, M. and I. Malitz, 1985. Investment patterns and financial leverage, in B. Friedman, ed., Corporate Capital Structure in the United States, (University of Chicago Press, Chicago, IL). Mandelker, G. and S. Rhee, 1984. The impact of the degrees of operating and financial leverage on systematic risk of common stock, Journal of Financial and Quantitative Analysis 19, 45-57.

L.G. Perry, J.NJ.N. . Rimbey/The Financial Review 3333 (1998) 85-98 9898M.A. M.AMoh'd, . Moh'd, L.G. Perry, Rimbey/The Financial Review (1998) 8598

9898

Mehran, H., 1992. Executive incentive plans, corporate control, and capital structure, Journal ofFinancial and Quantitative Analysis 27, 539-560. Myers, S., 1977. Determinants of corporate borrowing, Journal of Financial Economics 5, 147-176. Myers, S., 1984. The capital structure puzzle, Journal of Finance 39, 575-592. Myers, S. and N. Majluf, 1984. Corporate financing and investment decisions when firms have information that investors do not have, Journal of Financial Economics 13, 187-221. Parks, R., 1967. Efficient estimation of a system of regression equations when disturbances are both serially and contemporaneously correlated, Journal of the American Statistical Association 62, 500-509. Rhee, S., 1986. Stochastic demand and a decomposition of systematic risk, Research in Finance 6, 197-216. Rozeff, M., 1982. Growth, beta and agency costs as determinants of dividend payout ratios, Journal of Financial Research 5, 249-259. Scott, J., Jr., 1977. Bankruptcy, secured debt, and optimal capital structure, Journal of Finance 32, 1-20. Smith, C. and R. Watts, 1992. The investment opportunity set and corporate financing, dividend, and compensation policies, Journal of Financial Economics 32, 263-292. Titman, S. and R. Wessels, 1988. The determinants of capital structure choice, Journal of Finance 43, 1-19. Zimmerman, J., 1983. Taxes and firm size, Journal of Accounting and Economics 5, 119-149.